Estimating a game theoretic model

Applied game theory generally shows possible game situations, discussing their implications. This paper derives an estimation procedure for non-cooperative games, using cross-section data. During the transformation of the data an attempt is made to pick extreme points, to find the outer borders for the actual game. The estimation procedure is an algorithm that switches between several software packages. In an application a motivated choice can be made for the transformation functions such that the estimated game comes closest to the actual game of the empirical problem under study. To demonstrate the estimation procedure, a situation is studied in which a villager voluntarily chooses a level of participation in forest management.